Abstract
Crossing a busy intersection requires a rapid decision whether to go or not. Decision accuracy generally improves over time, with longer accumulation of visual motion information. However, the time available to accumulate sensory information may be constrained by the complexity and movement time of the required action. This factor is rarely accounted for in perceptual decision-making studies, which typically rely on simple button presses as indicators of decision outcome. Here we ask whether action-response modes (manual button press vs. interceptive hand movement) directly affect the accuracy and timing of a go/no-go decision. We recorded eye and hand movements of n=10 human observers while they viewed a small disc moving toward a goal. The target disappeared 100, 200, or 300 ms after onset and observers had to indicate whether the target would miss (no-go decision required) or pass (go decision required) the goal by either inhibiting a response, or by executing a button press or interceptive hand movement, respectively. Decision accuracy increased with increasing target presentation duration, confirming that longer sensory accumulation improves decision accuracy. Across presentation durations, decision accuracy was significantly higher and less variable when observers indicated their decision by pressing a button, compared to when they manually intercepted the target. To compensate for different movement execution times, observers initiated their hand movements ~270 ms earlier and intercepted ~70 ms later when manually intercepting the target compared to pressing the button. These results indicate that observers had less time to form their decisions when performing goal-directed hand movements compared when they simply had to press a button. We propose that the more complex planning and execution of the interceptive hand movement competes with decision formation. Our results highlight critical differences in sensorimotor decision processes between simple button press and more complex hand movement tasks.